For everyday research users: start from an Excel or LibreOffice table and work through data cleaning, statistical analysis, exporting results and preparing paper tables and figures.
Standardise field names, units, group information and sample IDs, and keep a backup of the raw data.
Don't overwrite the raw data; set up three folders: raw, clean and result.
Check for missing values, duplicate records, outliers, date formats and inconsistent category names.
Keep the cleaning process traceable so you can review it during revisions.
Import the cleaned data into statistics software and confirm variable types, groups and units are correct.
After importing, start with descriptive statistics and visual checks.
Choose t-tests, ANOVA, correlation, regression or non-parametric tests according to the experimental design.
The statistical method must match the experimental design; don't choose based on the menu alone.
Export statistical tables, significance annotations and figures for the results section.
Make sure units, sample sizes and significance notes in tables and figures are complete.
